Garrett: The ongoing fight against GPL enforcement

Posted Jan 31, 2012 15:54 UTC (Tue) by dwmw2 (subscriber, #2063)
Parent article: Garrett: The ongoing fight against GPL enforcement

Under British copyright law, at least until Temple Island vs. New English Teas gets overturned by someone with half a brain, any reimplementation of Busybox would probably still be considered to be an infringement of the original copyright of Busybox ☺

Garrett: The ongoing fight against GPL enforcement

Posted Jan 31, 2012 16:09 UTC (Tue) by karim (subscriber, #114) [Link] (2 responses)

Wow. Now that's something some software companies would love to have in the US.

Garrett: The ongoing fight against GPL enforcement

Posted Jan 31, 2012 19:51 UTC (Tue) by Wol (subscriber, #4433) [Link] (1 responses)

This has been discussed on Groklaw. Two points to bear in mind.

Firstly, the two companies were already in court because the defendant had copied, and used commercially, an image belonging to the plaintiff. This is a criminal offence (I know it isn't in the US, but it is over here).

Secondly, the defendant studied the original picture, and set out to produce a replacement. In literature, this would be a clear case of copyright violation - producing a derivative work without permission. I know "photography is different", but as somebody on GL pointed out, people who use clean-room re-implementation to avoid copyright violation make a point of *avoiding* intimate knowledge of the product they are copying. The defendants *studied* the product they were copying.

So that's why Judge Birss ruled the way he did. Anybody trying to build on this to claim "your photo is similar to mine" is likely to find themself on a sticky wicket.

Cheers,
Wol

Garrett: The ongoing fight against GPL enforcement

Posted Jan 31, 2012 20:12 UTC (Tue) by BrucePerens (guest, #2510) [Link]

CAI v. Altai isn't all that dissimilar, and is in U.S. courts, and is about software.

Garrett: The ongoing fight against GPL enforcement

Posted Jan 31, 2012 19:59 UTC (Tue) by landley (guest, #6789) [Link] (6 responses)

I still retain my original copyrights to all the code I contributed to busybox. The SFLC's right to represent those copyrights was terminated in 2008. They still have Erik Andersen and Denys Vlasenko on board, but I have standards documents and man pages to point to as implementing documented standards, and all my code's public if you'd like to compare for similarity.

Oh, and Bruce Fscking Perens got the idea of a swiss-army-knife executable from gzip, it was not original with him:

http://busybox.net/~landley/forensics.txt

I started toybox in 2006 in part because Bruce's GPLv3 trolling made busybox unpleasant to even _look_ at for me (http://lwn.net/Articles/202106/), but I still wanted to work on that _kind_ of thing because I felt I could do a BETTER JOB, from an engineering standpoint. Which means a radically different implementation.

I solved the _technical_ challenges, but then had to figure out whether or not to continue the project in context with busybox. I explicitly mentioned "undermining my hand-picked successor" on busybox as a reason NOT to continue:

http://landley.net/notes-2008.html#30-06-2008

And didn't... until the market vacuum on android was pointed out to me by Tim. Toybox was all my code, which I could BSD license if I wanted to, and I thought it over fairly extensively before doing so:

http://landley.net/notes-2011.html#13-11-2011

Years ago I wouldn't have considered doing so, but now?

http://landley.net/notes-2011.html#16-12-2011

Bring it.

Garrett: The ongoing fight against GPL enforcement

Posted Jan 31, 2012 20:20 UTC (Tue) by BrucePerens (guest, #2510) [Link] (3 responses)

Oh, and Bruce Fscking Perens got the idea of a swiss-army-knife executable from gzip, it was not original with him:

You've repeated this one for years, and it's wrong, wrong, wrong. I did not get the idea of linking all of the names to one executable from gzip, and I never claimed that particular element of busybox as an invention.

I first worked on Unix at the NYIT Computer Graphics Laboratory, the predecessor of Pixar, in 1981. At the time we had Version 6 Unix on PDP-11 and we were just getting the first VAX to be released from DEC. The device of linking multiple names to one executable was present in the command line tool set of Version 6 Unix, and you can probably go to a V6 source archive and find it there today. It was not invented as part of gzip.

Garrett: The ongoing fight against GPL enforcement

Posted Jan 31, 2012 20:25 UTC (Tue) by landley (guest, #6789) [Link] (2 responses)

I didn't say it was invented as part of busybox, I was pointing out an instance of it predating busybox in one of the components of the original busybox. I.E. here's proof it WASN'T an original invention in busybox.

Garrett: The ongoing fight against GPL enforcement

Posted Jan 31, 2012 20:41 UTC (Tue) by BrucePerens (guest, #2510) [Link]

This is the straw man fallacy. I never claimed to invent it, you argue that I didn't invent it.

What I did was make a tiny replacement of the entire Unix command line toolkit necessary for a limited purpose that ran like the command line programs it replaced, not like the "stand-alone shell" programs that people were making around then. I made the first, what follows are destined to be clones and copies.

Garrett: The ongoing fight against GPL enforcement

Posted Jan 31, 2012 20:45 UTC (Tue) by donbarry (guest, #10485) [Link]

For the record, the pnmtools set also used the method of swiss-army-knife linking by the early 90s.

Garrett: The ongoing fight against GPL enforcement

Posted Jan 31, 2012 20:47 UTC (Tue) by welinder (guest, #4699) [Link] (1 responses)

> Oh, and [...] Perens got the idea of a swiss-army-knife executable
> from gzip, it was not original with him.

That idea is older than BusyBox. Much, much older. I was doing that
in the mid-80s and I am certain it wasn't original.

Garrett: The ongoing fight against GPL enforcement

Posted Feb 1, 2012 1:21 UTC (Wed) by landley (guest, #6789) [Link]

Agreed. I misspoke: my point was that he didn't come up with it, and in fact included a component that was already using that trick in its original upstream version.

I didn't mean that nobody had ever used it before that.